[gpt5] Are verbosity and reasoning effort already supported within semantic kernel in python?

[lmatzi]

The release of GPT5 brings a few more settings on the plate like verbosity and reasoning in the responses feature of openai, e.g.
https://platform.openai.com/docs/guides/latest-model?quickstart-panels=fast


Are there already settings available in the semantic kernel connectors to utilize those 2 parameters verbosity and reasoningusing AzureChatCompletion?

Intuitively I would assume that the parameter resoning_effort in the AzureChatPromptExecutionSettings which allows ('low','medium','high') would be the one, but according to documentation 'minimal' would not be supported. Is this already available?

[moonbox3]

Hi @lmatzi the screenshots you posted are making Responses API calls. SK Python has support for the Responses API via the AzureResponsesAgent. You can find support for the reasoning effort here, along with reasoning summaries (if you have opted-in to that feature via Azure OpenAI): https://github.com/microsoft/semantic-kernel/blob/main/python/samples/concepts/agents/openai_responses/responses_agent_reasoning.py.